Sequence specific recruitment of proteins to the inner membrane complex of the malaria parasite Plasmodium falciparum (Welch, 1897)

(1)

Sequence specific recruitment of proteins to the inner membrane

complex of the malaria parasite

Plasmodium falciparum (Welch, 1897)

Dissertation

Zur Erlangung des Doktorgrades an der Fakultät für Mathematik,

Informatik und Naturwissenschaften,

Fachbereich Biologie

der Universität Hamburg

vorgelegt von

Johanna Wetzel

(2)

Date of oral defense: 19 December 2014

The following evaluators recommend the admission of the dissertation:

Prof. Dr. Tim-Wolf Gilberger Prof. Dr. Christian Voigt

(3)

I Abstract

Plasmodium falciparum causes malaria tropica, the most severe form of this disease. There

are about 250 to 500 million infections annually and about 600’000 deaths per year, the majority being children younger than five years in sub-Saharan Africa. Drug resistance is widespread and there is no vaccine available. The intra-erythrocytic developmental cycle of this parasite causes all symptoms associated with malaria. Plasmodium belongs to the infra-kingdom Alveolata, which consists of single-celled organisms that are unified by the existence of an endomembrane system that underlies the plasma membrane, termed “inner membrane complex “ (IMC). The IMC has distinct and crucial functions for the respective organisms in cytokinesis and host cell invasion. More than 30 structurally and phylogenetically distinct proteins are embedded in the IMC membranes, where a portion of these proteins displays N-terminal acylation motifs. While N-terminal myristoylation is catalyzed co-translationally within the cytoplasm of the parasite, palmitoylation takes place at specific membranes and is mediated by palmitoyl acyltransferases (PATs). Here, for the first time, a PAT (PfDHHC1) is identified that exclusively localizes to the IMC. It is shown that this enzyme has an identical distribution during schizogony like two putative substrates; the dual acylated IMC proteins (PfISP1 and PfISP3). Using a comprehensive mutagenesis approach, both proteins were used to probe into specific sequence requirements for IMC membrane recruitment, and their interaction with differentially localized PATs of the malaria parasite.

(10)

II Abbreviations

2BP 2-bromopalmitate

3xHA triple hemaglutinin

A alanine

aa amino acid

ABE acyl biotin exchange assay

ACT artemisinin combination therapy

AKR1/2 ankyrin repeat-containing protein 1/ 2

AMA1 apical membrane protein 1

Amp ampicillin

AP2 Apetala2

ApiAP2 apicomplexan AP2

APS ammonium persulphate

APT acyl protein thioesterase

ARO armadillo repeats only protein

ATP adenosine triphosphate

BiP binding immunoglobulin protein

BNI Bernhard-Nocht-Institute for Tropical Medicine

bp base pair

BMCC 1-Biotinamido-4-[4'-(maleimidomethyl)cyclohexane-

carboxamido]butane

BSA bovine serum albumin

BSD blasticidin deaminase

C cysteine

C- Carboxy-

cam calmodulin

cDNA complementary DNA

CD36 cluster of differentiation 36

CDPK1 calcium-dependent phosphokinase 1

ChAd63/MVA chimpanzee adenovirus 63 / modified vaccinia virus Ankara

(11)

D asparagine Da Dalton DAPI 4′,6-Diamidino-2-phenylindol DBD DNA-binding domain dd double distilled DDT dichlorodiphenyltrichloroethane DHHC asparagine-histidine-histidine-cysteine

DMSO dimethyl sulfoxide

DNA desoxyribonucleic acid

dNTP deoxyribonucleotide triphosphate mix

DRM detergent-resistant membrane

dsDNA double stranded DNA

DTT 4,4'‐(2,2,2‐trichloroethane‐1,1‐diyl)bis(chlorobenzene)

E glutamic acid

E. coli Escherichia coli

EBA erythrocyte binding antigen

EDTA ethylenediaminetetraacetic acid

ERD2 ER lumen protein retaining receptor 2

ER endoplasmatic reticulum

ERF2 effect on Ras function

EtBr ethidium bromide

EtOH ethanol

F phenylalanine

fw forward

G glycine

g 9.81 m/s2

gDNA genomic DNA

GAP genetically attenuated parasite

GAP45/50 glideosome-associated protein 45/50

GAPDH glyceraldehyde 3-phosphate dehydrogenase

GFP green fluorescent protein

GRASP Golgi re-assembly stacking protein

GTF general transcription factor

(12)

h hour

H3 histone 3

H4 histone 4

HA hydroxylamine

HCl hydrogen chloride

hDHFR human dihydrofolate reductase

HeLa cells Henrietta Lacks cells

HEPES N-2-hydroxyethylpiperazine-N-2 ethansulphonic acid

HP1 heterochromatin protein 1

hpi hours post invasion

HRP horseradish peroxides

I isloleucine

ICAM1 intracellular adhesion molecule 1

IDC intra-erythrocytic developmental cycle

IFA immuno fluorescence assay

IgG immunoglobulin G

IMC inner membrane complex

IP immunoprecipitation

IPTG isopropyl-β-D-thiogalaktopyranosid

iRBC infected red blood cell

ISP IMC sub-compartment protein

K lysine

kb kilobase

kDa kilo dalton

L leucine

l litre

LB luria broth

M molar, methionine

MBOAT membrane-bound -acyl transferase

mCherry Cherry fluorescent protein

MCS multiple cloning site

MDCL Michael G. DeGroote Center for Learning and Discovery

(13)

ml millilitre

MLC myosin light chain

mM millimolar

mRNA messenger RNA

MSP merozoite surface protein

MTIP myosin tail interacting protein

MW molecular weight

MyoA myosin A

Myr-CoA myristoyl-CoA

N asparagines

N- Amino-

NaAc sodium acetate

NaCl sodium chloride

NEM N-ethylmaleimide

ng nanogram

NMT N-myristoyl transferase

NP-40 nonidet P-40

OD optical density

ORF open reading frame

P proline

P. falciparum Plasmodium falciparum

PAT palmitoyl acyl transferase

PBS phosphate-buffered saline

PBS-T PBS-Tween

PCR polymerase chain reaction

pDNA plasmid DNA

PFA4/5 protein fatty acyltransferase

PfEMP1 Plasmodium falciparum erythrocyte membrane protein

PM plasma membrane

PMSF phenylmethylsulfonyl fluoride

PPT protein palmitoyl thioesterase

PV parasitophorous vacuole

PVM parasitophorous vacuole membrane

(14)

R arginine

RALP rhoptry associated leucine-rich protein

RAP rhoptry associated protein

RAS radiation attenuated sporozoite

RBC red blood cell (erythrocyte)

RESA ring-infected erythrocyte surface antigen

RNA ribonucleic acid

RON1 rhoptry neck protein 1

ROP rhoptry protein

rpm rounds per minute

RPMI Rosswell Park Memorial Institute

RT room temperature

rv reverse

S serine

S. cervisiae Saccharomyces cerevisiae

sec seconds

SDS sodium dodecyl sulfate

SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis

SP signal peptide

STET sucrose, Tx-100, EDTA, Tris

Swiss TPH Swiss Tropical and Public Health Institute

SWF1 spore wall formation protein 1

T threonine

TAE tris acetate

TCA trichloroacetic acid

TE tris-EDTA

TEMED tetramethylethylenediamine

TF transcription factor

TMD transmembrane domain

TRAP trombospondin related adhesion protein

TVN turbovesicular network

(15)

UAS upstream activation sequence

5’ UTR 5’ untranslated region

UV ultraviolet

V volt, valine

VAC8 vacuole related protein

Y tyrosine

W tryptophan

WHO world health organization

WR WR99210

µF microfarad

µg microgram

(16)

1. Introduction

The causative agent of malaria is an obligate intracellular apicomplexan parasite of the genus

Plasmodium. Among over 200 Plasmodium species, only five of them are pathogenic to

humans: P. falciparum and P. vivax, but P. ovale, P. malariae and the monkey malaria P.

knowlesi can cause infections (Singh et al., 2004). Plasmodium falciparum causes malaria

tropica (also termed falciparum malaria), the most severe form of human malaria that is responsible for 80 % of all cases (Mendis et al., 2001). Plasmodium falciparum is most prevalent in Africa, whereas P. vivax is mainly present in Southeast Asia. In other parts of Asia and South America, prevalences of P. falciparum and P. vivax are equal, since transmission rates are lower.

Malaria tropica represents one of the major human health problems in endemic countries and also causes an economical and social burden (Sachs and Malaney, 2002). There are three billion people at risk, 150 to 300 million people are affected annually and about 600’000 die per year due to malaria (WHO World Malaria Report, 2013). Children under five years in sub-Saharan Africa are mostly affected, representing 80 % of all cases (Snow et al., 2005). This disease is widespread among subtropical and tropical countries, and its highest burden is in Africa, but as well in Southeast Asia and South America (Fig. 1.1). Poverty has been and still remains a major reason for the disease (Sachs and Malaney, 2002). Transmission of parasites takes place by Anopheles mosquitoes, which also represent the definitive hosts where the sexual reproduction takes place. Anopheles gambiae is a major vector since it is especially robust and effective, however 41 of the more than 400 species of the family of anopheline mosquitoes are dominant vectors for malaria transmission to humans (Hay et al., 2010). A recently established global map shows the highly complex variety of Anopheles species distribution (Sinka et al., 2012): the situation in Africa is characterized by co-dominance of fewer species compared to the Asian-Pacific or Central-America, where numerous species co-exist (Sinka et al., 2012).

The lack of an effective vaccine, spreading resistances of the parasite to antimalarial chemotherapeutics and problematic vector control measurements are the basis for the urgent need for new intervention approaches in the fight against this disease.

(17)

Figure 1.1 World map representing P. falciparum prevalence. Nowadays, malaria is endemic to

over 100 nations, mostly affected are tropical and subtropical countries. Highly endemic regions are indicated in red, less endemic regions are in lighter shades of red. Image was adapted from sanger.ac.uk.

Figure 1.1 World map representing P. falciparum prevalence.

1.1 Symptoms, treatment and prevention of malaria

The word “malaria” originates form the Italian words “mala” (bad) and “aria” (air). It was believed that foul gasses released from the soil and water of the swamps cause the disease. This idea persisted throughout the 19th century until Laveran discovered the parasite (Laveran, 1880).

1.1.1 Symptoms

It generally takes about two weeks (Bartoloni and Zammarchi, 2012) for humans to develop malaria symptoms after infection, but the incubation period can range from 9 to 30 days. Typically, individuals suffer from paroxysmal attacks – shivering and coldness follow periods of fever and sweating in characteristic intervals. The attacks usually consist of a phase of coldness where the surroundings are perceived as extremely cold even though the body temperature is steadily increasing. Then, a phase of extreme heat arises where patients feel extremely hot and sick for several hours, but without sweating. Temperatures of more than 40° C can be reached. Subsequently, the body temperature is slowly going back to normal, while patients go through a period of extreme sweating. According to the timing of the fever

(18)

intervals malaria infections are categorized. Malaria quartana (or quartan fever) is caused by

P. malariae and occurs every three days with fever on the first and forth day with 2 days

without fever in between. Malaria tertiana (tertian fever) occurs every two days and is caused by P. vivax or P. ovale. Paroxysms are less prominent in P. falciparum infections (malaria tropica), since the parasites grow asynchronously with mostly uninterrupted fever and sometimes waves of fever every 48 hours (White, 2013).

The WHO categorizes malaria cases into “uncomplicated” and “severe”, based on the severity of symptoms. Common symptoms of uncomplicated malaria include fever, headaches, chills, vomiting, muscle pains and anemia. However, uncomplicated malaria cases can develop into severe malaria, which is characterized by symptoms that are much more serious such as acute renal failure, severe anemia and respiratory failure that can lead to coma and death. Mortality increases when the parasitaemia of an infected individual is greater than 2 %. Most severe malaria cases occur in non-immune individuals that acquire malaria tropica (White, 2013). This is mostly due to the fact that the parasite exports P. falciparum erythrocyte membrane protein 1 (PfEMP1) to the erythrocyte surface, which mediate sequestration of the parasites within the capillaries, which has particularly severe effects in organs and the peripheral bloodstream (see below, 1.2.2; Kayser et al., 2005; Kyes et al., 2007).Non-falciparum severe malaria is rare, but has been observed in individuals infected with P. vivax. In these cases, death is usually caused by severe anemia, renal failure or coma. Cerebral malaria is a form of severe malaria that shows further neurological symptoms like seizures and coma. Particularly infants are prone to severe malaria, since they are very vulnerable and semi-immunity has not built up yet. Older individuals living in high-transmission areas usually develop semi-immunity over the years, protecting them against severe malaria, but not against infection or mild symptoms. This semi-immunity cannot be transferred through pregnancy and can only be acquired over time after exposure to different strains, although complete protection from disease will never be achieved (White, 2013).

1.1.2 Treatment

European settlers and their slaves brought malaria to the Americas in the 16th century, but the earliest treatment was discovered shortly afterwards. Quinine, which is found in the bark of the Peruvian Cinchona tree, was recognized to be very effective and the dried powder was usually consumed, mixed with red wine. This natural compound is still comparably effective,

(19)

available, although the parasite developed resistance against all of them (Dondrop et al., 2009; Phyo et al., 2012).

Besides the natural compounds described above, there are several classes of synthetic compounds used for malaria treatment (Fidock et al., 2004; Greenwood et al., 2008; Delves et al., 2012). The most prominent class, the 4-aminoquinolines (including chloroquine, amodiaquine and piperaquine), consists of usually fast-acting drugs (Greenwood et al., 2008). Moreover, the 8-aminoquinolines (e.g. primaquine), and the 4-methanol-quinolines (e.g. quinine, quinidine, halofantrine and mefloquine) and antibiotics (such as azithromycin, doxycycline, tetracycline and clindamycine) are effective to treat malaria (Draper et al., 2013). Antifolates such as atovaquone, pyrimethamine, sulphadoxine and proguanil are schizonticides and are also often used for treatment (Delves et al., 2012).

The natural compound artemisinin and its derivatives are the drug of choice since the year 2000, because it is very potent and effective, especially in combination with other antimalarials with longer half-life, known as artemisinin combination therapy (ACT). Even though resistance against artemisin developed 5 years ago, it is still effective in ACT (Dondrop et al., 2009; Phyo et al., 2012).

First choice for treatment consists of administration of ACT: I) coartem (artemether and lumefantrine), II) artesunate and amodiaquine, III) artesunate and sulphadoxin/pyrimethamine (SP), intravenously) SP and amodiaquine. Severe malaria is usually treated with artesunate, artemether or quinine. Vivax malaria can still be successfully treated with chloroquine. SP is the preferred drug to cure chloroquine-resistant malaria and to treat pregnant women.

Since resistance developed in 2008 at the border of Western Cambodia and Thailand-Myanmar (Noedl et al., 2008), new antimalarials are needed to treat the disease. One promising new candidate is the spiroindolone NITD609, which completely cures mice at a single oral dose of 100 mg/kg and an IC50 of 1 nM (Rottmann et al., 2010). Furthermore, NITD609 is the second gametocytocidal drug discovered besides primaquine and is currently undergoing phase IIa clinical trials (van Pelt-Koops et al., 2012).

(20)

1.1.3 Prevention

Vector control and chemoprophylaxis

Attempts to prevent malaria infection by controlling the vector have been reasonably successful over the past years. Among only four classes of insecticides, dichlorodiphenyltrichloroethane (DDT) and pyrethroids are the most well known. It was once widely used to kill mosquitoes, but numerous reports informed about toxic effects on humans and other animals, as well as being an endocrine disruptor, which led to a stop of use in 1972 (Turusov et al., 2002). Between 1947 and 1952, malaria was eliminated from the North America by spraying DDT. Nowadays, DDT is still used in endemic countries as a chemical for indoor residual spraying (Enayati et al., 2010; Pluess et al., 2010). In contrast to DDT, pyrethroids are safer for vertebrates, but toxic to insects and invertebrates, and are a widespread component of household insecticides (Maund et al., 2012). However, the mosquito vector developed resistance against both chemicals by mutations of the sodium channel (Hemingway et al., 2004; Enayati and Hemingway, 2010). Furthermore, mosquitoes developed metabolic resistance by overexpressing glutathione-S-transerase, which confers resistance through detoxification of both chemicals (Enayati and Hemingway, 2010).

Especially in high-transmission areas, the combination of a physical biting barrier that consist of insecticide-treated mosquito nets and indoor residual spraying have been shown to be highly effective (Lengeler, 2004; Pluess et al., 2010). Their use has increased significantly over the last years, since several organizations expanded the distributions over the last years. It was reported that up to 73 % of the households in sub-Saharan Africa use bed nets, which evidently contributed to reduction of malaria prevalence in recent years (WHO, 2013). Use of insect repellents is another method to protect oneself from mosquito bites, but it is not very feasible for people living in endemic regions due to the high cost. It is, however, rather suitable for travelers. Removal of open water, the breeding ground for mosquito larvae is key to reduce transmission and decrease Anopheles density in rural areas. Furthermore, health education is needed in order to recognize early symptoms.

Additionally, many of the available antimalarials can also be used as chemoprophylaxis. Once again, this method is only suitable for travelers, because of severe side effects and high costs. Depending on the area and endemicity, there are different recommendations for chemoprophylaxis. In case of uncertainty, malarone (atovaquone-proguanil), doxycycline, primaquine or mefloquine are recommended.

(21)

Towards the development of a malaria vaccine – RTS,S

The low level of commercial interest, the complex, mostly intracellular life cycle of the parasite, its antigenic variation of surface proteins, large gaps in understanding Plasmodium biology and its interaction with the human immune system are some reasons why a malaria vaccine is not yet available (Gardner et al., 2002; Florens et al., 2002).

Besides all these adversities, the first malaria vaccine called RTS,S is expected to be available for the public next year (Hill, 2011). RTS,S is a subunit vaccine, consisting of the hepatitis B surface antigen fused to the central repeat region and thrombospondin domain of the circumsporozoite protein (CSP), the main surface protein of sporozoites. It is the most advanced malaria vaccine currently in development. However, it seems to be a “leaky” vaccine, which protects only against a certain number of bites and seems to be especially unreliable in hyperendemic regions. The phase III clinical trial conducted from 2009 to 2012 in 5 to 17 months olds revealed a 56 % reduction in acquisition of clinical malaria and a 47 % reduction of progression into severe malaria 12 months after vaccination (Agnadji et al., 2011; RTS,S Clinical Trials Partnership 2012). These numbers even further decreased after 18 months, when protection against malaria was 46 % and only 36 % for protection against severe malaria.

The Malaria Vaccine Technology Roadmap, funded by the Bill and Melinda Gates Foundation, aims to accelerate the licensing of a malaria vaccine. It is the aim to produce a vaccine that is 50 % protective against severe malaria by 2015 and to further achieve an efficacy of 80 % by 2025 (Hill, 2011; Malaria Vaccine Technology Roadmap, 2013). However, this depends on the outcome of the results from phase III clinical trials in late 2014 (Malaria Vaccine Technology Roadmap, 2013).

WPV – whole parasite-based vaccine

The idea to produce a vaccine based on attenuated, whole parasites was first realized about 50 years ago (Nussenzweig et al., 1967). They showed that vaccination with radiation-attenuated sporozoites (RAS) protect mice from malaria infection and were able to show protection in humans (Nussenzweig and Nussenzweig, 1989). Further studies investigating this method showed that the vaccine is well-tolerated and safe, but showed a relatively low level of protection, probably due to the high diversity of Plasmodium strains in the field. When administered by mosquito bite, the vaccine showed a higher level of protection (Khan et al.,

(22)

2012). The vaccine shows sterile immunity in animals and almost complete protection in humans when administered by bite, although more than 1000 mosquito bites are needed to achieve this result. A major concern is to produce a balanced level of radiation dose to ensure DNA damage without reducing infectivity and immunogenicity of the sporozoites too much (Ménard, 2013). The production of high numbers of infective sporozoites is commercialized by the company SANARIA and clinical trials using cryopreserved sporozoites injection are ongoing (Seder et al., 2013; Shekalaghe et al., 2014).

More recently, this old idea has gained more attention, but a modern twist has been put to it: genetically attenuated sporozoites (GAPs) harbor gene deletion, which leads to an arrest of parasite growth at the late liver stage. In rodents, this approach seems promising and GAPs have shown a higher potency in inducing protective immunity than radiation-attenuated sporozoites (Butler et al., 2012). Those parasites are genetically homogeneous and defined and preliminary data indicated that GAPs are highly efficacious in producing sterile protection in mice (Labaied et al., 2007; Khan et al., 20012; Butler et al., 2012). Production of

P. falciparum late-arresting GAPs and their evaluation in humans showed some potential

(VanBuskirk et al., 2009; Epstein et al., 2011). However, there was one breakthrough parasite that caused disease, even though the genome still contained the respective gene deletions (Spring et al., 2013). Clinical trials comparing these new methods are needed to evaluate and to directly compare them and to further test the efficiency in highly endemic regions.

Future directions, control and elimination

Low-transmission areas with seasonal malaria episodes have great potential for future malaria elimination. Spreading resistance to artemisinin and its derivatives threatens global malaria control, which makes monitoring and containment of the spread of resistance necessary. Mass drug administration might be a possible measure to contain resistance and eliminate malaria from hypoendemic regions.

(23)

1.2 Biology and life cycle of the parasite

1.2.1 Life cycle

Parasites are transmitted via the bites of infected female Anopheles mosquitoes, which necessitate blood meals for their oogenesis. In the human host, the injected sporozoites migrate through the blood stream to the liver, where they infect hepatocytes. During the following 1-2 weeks, the intra-hepatic liver-stage parasites multiply asexually and differentiate via schizogony into merosomes that contain merozoites (Prudêncio et al., 2006). This stage is called exo-erythrocytic schizogony whereas schizogony refers to nuclear divisions without cytoplasmic division. This development takes usually 6 (P. falciparum) to 15 (P. malariae) days (Collins and Jeffery, 2007; White, 2013). From the host cells emerge merosomes, which then rupture and release thousands of merozoites into the blood stream, where they infect erythrocytes (Ménard, 2013). Within the erythrocytes, the merozoites grow first into a ring-shaped form and develop further into trophozoites. At the subsequent schizont stage, the parasite nuclei divide several times within a common cytoplasm. The parasite feeds by ingesting hemoglobin, which is metabolized into amino acids in the food vacuole. The toxic by-product α-hematin (ferriprotoporphyrin IX) is stored as a biologically inert hemozoin crystal. Late schizonts are called segmenters, in which new merozoites develop after completion of cytokinesis that finally leave the red blood cells and invade additional erythrocytes (Fig. 1.2). This asexual intra-erythrocytic developmental cycle (IDC) is completed within 48 hours (in P. malariae within 72 hours) and causes all symptoms associated with malaria. Such massive proliferation can lead to a parasite load of 100 billion in the blood of an adult.

Besides schizogony a small percentage of parasites differentiate into sexual forms, male and female gametocytes (micro- and macrogametocytes) (Dixon et al., 2008). Gametocytes do not cause pathology in the human host and will disappear from the circulation after 3 weeks (P.

falciparum) if not taken up by a female anopheline mosquito (Drakeley et al., 2006).

However, these numbers can differ (reviewed in Drakeley et al., 2006). In the mosquito’s midgut, the gametocytes develop into gametes and fertilize each other, forming motile ookinetes. The resulting ookinete traverses the mosquito gut wall and encysts on the exterior of the gut wall as an oocyst. Here, they undergo several divisions producing a large number of sporozoites that cross the basal lamina into the mosquito’s body cavity and migrate to the

(24)

salivary glands. During the next blood meal, the sporozoites are injected into the subcutaneous tissue of the human host (Fig. 1.2). A proportion is ingested by macrophages and others are taken up by the lymphatic system where they are presumably destroyed (Amino et al., 2006). The sporozoites, which successfully enter the blood stream, move to the liver where they begin the cycle again.

Figure 1.2 Overview over the life cycle of Plasmodium between the human host and the anopheline vector. From an anthropocentric point of view, the most important part of the parasite’s

complex life cycle is the asexual, intra-erythrocytic, developmental cycle, which causes all of the symptoms associated with malaria. It is initiated by the invasion of the human red blood cell. Adapted from Ménard, 2013.

1.2.2 Species-specific aspects of parasite biology

Blood stage schizogony in P. falciparum differs from that of the other human malaria parasites with respect to the trophozoite- and schizont-infected erythrocytes, which adhere to

(25)

are also able to adhere to uninfected erythrocytes, so-called “rosetting”. PfEMP1 surface molecules that are encoded by the var gene family are known to mediate this sequestration (Kyes et al., 2007). The var multigene family consists of 60 members, which undergo antigenic variation, regulated by mutually exclusive expression of only one gene at a time (Scherf et al., 1998). PfEMP1 attaches to different endothelial receptors in capillaries, like ICAM1 in the brain, chondroitin sulphate A in the placenta or CD36 in many other organs. As a consequence, only early ring stage parasites circulate in the bloodstream. On one side, sequestration is an advantage for the parasite, because it prevents elimination of infected red blood cells (iRBCs) by the spleen. On the other side, sequestration interferes with blood circulation in vital organs and causes severe complications such as cerebral and placental malaria. Additionally, P. falciparum infects all erythrocyte stages and can, therefore, reach a higher parasitaemia than other Plasmodium species (Kayser et al., 2005).

During the development of the P. ovale and P. vivax parasites in the liver, so-called hypnozoites can emerge from sporozoites. These small, mononuclear stages may persist in hepatocytes for several months or even years and develop into merosomes in numerous episodes. In these cases, malaria-relapses are due to reactivation of hypnozoites, or due to recrudescence of persisting erythrocytic stages (Kayser et al., 2005).

1.2.3 The P. falciparum genome

The haploid genome of P. falciparum has a size of 22.8 megabases and consists of 14 linear chromosomes. The mitochondrial genome has a size of approximately 6 kilobases (kb), the circular apicoplast genome has a size of 35 kb. The genome sequencing project revealed that

P. falciparum contains 5409 open reading frames (ORFs), of which 60 % code for

hypothetical proteins with an unknown function and no orthologous proteins in other eukaryotic species (Gardner et al., 2002; Aravind et al., 2003). With an AT-richness of 80.6 %, the genome of P. falciparum is the most AT-rich genome sequenced to date (Gardner et al., 2002).

(26)

1.2.4 Molecular architecture of the merozoite DNA containing and secretory organelles

Not only their highly adapted and extraordinary life cycle and immune evasion strategies make Plasmodium such an interesting study subject – their evolutionary origin and cellular architecture display many features that are not found outside the phylum (Cowman and Crabb, 2006). The most remarkable organelle is the apicoplast, which was acquired through secondary endosymbiosis (endo = within; syn = with; biosis = living) during evolution (Keeling, 2009; Van Dooren and Striepen, 2013). This ancient plastid-like organelle has 4 membranes surrounding the 35 kb genome, part of which is integrated into the parasite’s nuclear genome (Wilson et al., 1996). The apicoplast’s functions include fatty acid biosynthesis and it is additionally involved in major biochemical pathways (Yeh and DeRisis, 2011). The mitochondrial genome, in contrast, is only 6 kb small and lacks the characteristic cristae (van Dooren et al., 2005). Not all apicomplexan parasites have an apicoplast (e.g.

Cryptosporidium; Abrahamsen et al., 2004), but they all share the presence of an apical

complex. This aggregation of secretory organelles is important for invasion and consists of micronemes, rhoptries and dense granules. Micronemes emerge during schizogony from the Golgi complex and translocate alongside the microtubules and dock with the rhoptry tips (Bannister et al., 2003). Although it is still unknown how many different proteins are stored, some of the adhesines the parasite is using during erythrocyte invasion are stored in these organelles. The former include members of the erythrocyte binding antigens (EBAs), apical membrane antigen 1 (AMA1), and the trombospondin-related adhesion proteins (TRAPs). All of these proteins are targeting surface structures of the host cell (Malpede and Tolia, 2014). Compared to micronemes, the rhoptries are much bigger in size and less in numbers. Like micronemes they are Golgi-derived organelles and their protein contents are also involved in invasion, e.g. the family of reticulocyte-binding homologues (RH proteins) and the rhoptry neck proteins (RONs), but some also promote formation of the parasitophorous vacuole (PV) (Cowman et al., 2012; Counihan et al., 2013). The dense granule proteins function later during invasion and modify the host cell, e.g. RESAs. Rhoptries and micronemes localize apically, whereas dense granules are distributed throughout the cell’s cytosol.

(27)

Figure 1.3 Schematic representation of the ultrastructure of the merozoite. The merozoite is

with 1 to 2 µm the smallest life cycle stage and also one of the smallest eukaryotic cells in general. It is equipped with a full set of organelles, including micronemes, rhoptries and dense granules. These secretory organelles contain proteins that are crucial for invasion. In addition to the phylum-specific organelles (micronemes, rhoptries, dense granules, apicoplast), the parasite possesses all typical eukaryotic organelles. Cellular features are not according to scale. Schematic was adapted from Bannister et al., 2003.

The cytoskeleton

In general, membrane skeletons are required for cell shape, strength and maintenance, internal organization of subpellicular organelles, cell division and movement. They consist of filamentous protein networks that are linked to membranes through interactions with integral membrane proteins. Apicomplexan parasites need a robust but also dynamic cytoskeletal architecture to maintain structural integrity during the fast intra-erythrocytic growth phase and especially during host cell invasion. As described above, during the late phase of schizogony the parasite builds up its invasion machinery. Beside the secretory organelles, a highly complex endomembrane system is synthesized that lies underneath the merozoite’s plasma membrane (PM) in mature merozoites. This membrane system is termed the inner membrane complex (IMC, please refer to section 1.3) (Dubremetz and Torpier, 1978; Morrissette et al.,

(28)

1997; Raibaud et al., 2001; Morrissette and Sibley, 2002; Kono et al., 2012). PM and IMC are jointly form a triple bilayer called the pellicle. Below the PM and the IMC lie 2-3 sub-pellicular microtubules. Together, the subsub-pellicular microtubules and the IMC form the subpellicular network, which also contains interwoven 8-10 nm filaments that provide the cell with strength and stability. The subpellicular microtubules are organized form the apical polar ring (Bannister et al., 2003). The filaments of the SPN are linked together by intermembrane particles (IMPs).

The motor unit: The “glideosome”

Parasite motility is referred to as “gliding motility”. It is defined as an amoeboid-like movement of sporozoites (Steward and Vanderberg, 1988; Morrissette and Sibley, 2002; Baum et al., 2006). Extracellular adhesins from the thrombospondin-related anonymous protein (TRAP) family are linked to the IMC via the motor complex (Bullen et al., 2009). This sophisticated actin-myosin machinery is termed “the glideosome” (Fig. 1.4). Some of the characterized protein elements are the glideosome-associated proteins (GAPs; GAP45 and GAP50) that in association with myosin A (MyoA) and myosin light chain 1 (MLC1) (Cowman et al., 2012) are linked to the outer membrane of the IMC, potentially by interacting with transmembrane proteins of the IMC (Fig. 1.4) (Mann and Beckers, 2001; Gaskins et al., 2004; Baum et al., 2008; Bullen et al., 2009; Rayavara et al., 2009). The N-terminal domain of MLC1 serves as a tail for MyoA and brings the motor to its site of action by association with the C-terminus of GAP45. Depletion of GAP45 results in impaired gliding motility, invasion and egress in Toxoplasma gondii (Frénal et al., 2010). GAP45 has been implicated in the recruitment of the motor complex as well as in the maintenance of pellicle cohesion (Frénal et al., 2010). The N-terminus of GAP45 has acyl modification sites and mutational analysis indicated that these are essential for the insertion of the protein into the IMC (Rees-Channer et al., 2006; Frénal et al., 2010).

(29)

Figure 1.4 Schematic of the glideosome complex. This model visualizes some of the currently

known glideosome proteins and their interaction with each other. This schematic was adapted from Frénal et al., 2010.

1.2.5 Merozoite invasion of erythrocytes

After rupture of the host cell, it takes less than 30 seconds for a released merozoite to invade a new erythrocyte (Treeck et al., 2009). After encounter of an erythrocyte, it reorients on the surface of the erythrocyte to position its apical tip in direct contact with the targeted membrane and subsequently invades (Fig. 1.5). This invasion is crucial for the parasite’s survival and the invasion process is therefore generally accepted as a target for vaccine or drug development. Even though the merozoite is capable of invading a new host cell very quickly – speed alone is not enough. The cell further harbors several immune-evasion mechanisms that make it so successful. The proteinaceous surface coat of the merozoite is characterized through the presence of the merozoite surface protein 1 (MSP1), being the most abundant one. Different variants of MSP1 exist, indeed, with its particularly high level of polymorphism it is one of the most highly polymorphic eukaryotic proteins known (Volkmann et al., 2002). MSPs mediate the initial reversible attachment followed by a reorientation (Cowman and Crabb, 2006). The latter process is carried out by proteins of the EBA and RH protein families, which bind the red blood cell tightly (Wright and Rayner, 2014). One key player of the invasion process is the microneme protein apical membrane antigen 1 (AMA1) that is released onto the parasite’s surface just after egress of the merozoite from the iRBC. AMA1 forms a complex with the RONs, which are secreted from the rhoptries onto the iRBC surface (Lamarque et al., 2011). The newly formed molecular seal, the so-called tight junction, allows other rhoptry contents to get passed into the erythrocyte

(30)

(Fig. 1.5) (Cowman et al., 2012). Active invasion of the parasite is then powered by the “glideosome” (please refer to section 1.2.4). Merozoite surface proteins are shed off and the parasitophorous vacuole starts forming. The sealing of the vacuole and dense granule secretion marks the end of invasion. Although about 50 proteins are implicated in invasion (Cowman et al, 2012), the invasion-related processes might be mediated by a complicated protein network. Using transcriptional profiling approach in combination with phylogenetic profiles, domain-domain interactions and yeast two-hybrid datasets a high confidence protein network was constructed that governs invasion (Hu et al., 2010). This sub-network harbors about 400 proteins, wherefrom 260 are unknown hypothetical proteins.

Figure 1.5 Schematic of the different steps during merozoite invasion. A. Initial recognition is

followed by B. reorientation and tight junction formation. C,D. Active invasion takes place while the tight junction (brown) is moving towards the basal end of the merozoite while the surface coat is shed.

E. After complete invasion, the parasite starts growing and remodeling its host cell. Adapted from

Cowman et al., 2012.

1.3 The inner membrane complex (IMC)

1.3.1 Evolution of the IMC

While the secretory organelles are characteristic of all organisms belonging to the apicomplexan phylum (Levine; 1980), the IMC is a morphological trait of a large phylogenetic group called Alveolata (Adl et al., 2005). This phylogenetic group comprises three traditional main phyla: I) Dinoflagellata being typically marine flagellates; II) Ciliata

(31)

mostly parasitic species, including the genera of Theileria, Eimeria and Toxoplasma, which are important human and animal pathogens with significant impact (Wolters, 1991; Adl and Leander, 2007). A fourth group has been identified recently – the Chromerida consisting of marine, photosynthetically active protozoa (Moore et al., 2008). An exceptional characteristic of alveolates is the possession of the IMC also called alveoli (Cavalier-Smith, 1993; Gould et al., 2002; Adl et al., 2005; Bullen et al., 2009; Kono et al., 2012). This endomembrane system appears to possess taxon-specific functionality, exemplified above in the invasion process of the apicomplexan parasites. In dinoflagellates and ciliates the alveoli predominantly play a structural role. In all phylogentic clades it might also have a role as an important scaffolding element during cytokinesis (Gaskins et al., 2004; Sibley, 2004; Soldati et al., 2004).

1.3.2 Protein composition of the IMC

As a reflection of its multifunctional roles the IMC is composed of a phylogenetic, function and structural diverse set of proteins (Kono et al., 2012, 2013). They can be categorized into transmembrane proteins (like the glideosome-associated protein with multiple membrane spans 2, GAPM2), small acylated proteins (IMC sub-compartment proteins, PfISPs), alveolins (Alveolin 5; PF10_0039), alveolin like proteins (PF08_0033) and others (MAL13P1.228), which do not fit in any of the categories (Kono et al., 2013). A unique multigene family of proteins, the alveolins (Mann and Beckers, 2001; Gould et al., 2002; Gubbels et al., 2006; Gould et al., 2008; Bullen et al., 2009), is restricted to alveolates and, therefore, recognized as one molecular nexus of the Alveolata (Gould et al., 2008). So far, using proteomics (Gould et al., 2011), system biological approaches (Hu et al., 2010) and phylogenetic profiling (Kono et al., 2012) 28 IMC proteins have been identified. In addition to a common core set of conserved proteins, the IMC includes many lineage-specific proteins reflecting additional specialized roles. For instance, MAL13P1.228 is a Plasmodium-specific protein and is not found in any other species (Kono et al., 2012). Only a few IMC proteins are functionally characterized, most of them belong to the glideosome complex. Besides the GAPs, other peripheral IMC membrane proteins were identified. The ISPs belong to the group of small acylated IMC proteins. They were originally identified in Plasmodium (Hu et al., 2010) and described in greater detail in T. gondii (Beck et al., 2010; Fung et al., 2012). The IMC in T. gondii tachyzoites consists of 3 sub-compartments, the basal, central and apical part. TgISP1 is localized to the apical cap, TgISP2 and TgISP4 are localized centrally in the tachyzoite, and, ISP3 is localized to central and basal regions (Beck et al., 2010; Fung

(32)

et al., 2012). TgISP1 was found to have a gate-keeping function as it excludes the localization of other TgISPs to the apical cap (Beck et al., 2010). The gate-keeping function of TgISP1 may be accomplished by its C-terminal domain, because the C-terminal domain of TgISP2 could not complement it (Beck et al., 2010). TgISP2 seems to be the only vital ISP in T.

gondii, disruption of TgISP2 leads to defects in cell division (Beck et al., 2010). This data

suggests that this family of proteins, in contrast to GAP45, plays a role in cytokinesis. Some structural insights were delivered by recent work from Tonkin and colleagues, which revealed that TgISP1 possesses a small core domain with several cysteines (Tonkin et al., 2012). Moreover, crystallization of TgISP1 and TgISP3 showed that both proteins have a pleckstrin homology (PH) domain (Tonkin et al., 2014). PH domains mediate lipid and protein-protein interactions. Mutational analysis of selected amino acids of the PH fold and biochemical characterization implicate that TgISP PH domains do not bind phospholipids, but possibly function in binding proteins (Tonkin et al., 2014). The P. falciparum homologues PfISP1 (PF10_0107) and PfISP3 (PF14_0578) also possess N-terminal acyl modification motifs (Hu et al., 2010; Cabrera et al., 2012), but the sequence homology compared to the TgISPs is relatively low. However, both PfISPs contain a PH fold (predicted by SMART; Schultz et al., 1998; Letunic et al., 2012), indicating that PfISPs might play an important role in recruiting interaction partners to the IMC, which might fulfill a so far uninvestigated function.

1.3.3 IMC biogenesis

The duplication of the Golgi is among the earliest visible events of parasite replication and shortly precedes the onset of IMC biogenesis (Hu et al., 2002). The establishment of a polarized secretory system might be a prerequisite for daughter cell formation of the daughter cell IMC. The IMC is derived from clathrin-coated vesicles of the ER-Golgi complex (Bannister et al., 2000; Gordon et al., 2008; Yeoman et al., 2011). During Plasmodium schizogony the nuclei move to the periphery of the mother cell. Daughter cell formation is initiated by the formation of the IMC beneath the mother cell’s PM and an intact pellicle forms while the daughter cells grow about six hours before merozoite egress (Kono et al., 2012). In early schizonts, the IMC is visible as two cramp-like structures per nucleus (T1). They form while the mitotic spindles are still present. They further extend into small ring-like structures, about 630 nm in diameter (T2). Towards the end of schizogony the rings quickly

(33)

this typical dynamic are called “type A” proteins (Kono et al., 2012). In addition, a second IMC structure becomes visible about 3.5 hours prior egress, proteins showing this distinct localization pattern are called “group B” proteins. This group includes the alveolins (e.g. PF10_0039, PFE1285w) and the Plasmodium-specific MAL13P1.228. The later appearance of the group B proteins points towards a highly regulated mechanism through which proteins are recruited to the IMC. Moreover, the two different groups likely fulfill distinct functions.

IMC in gametocytes

The pre-sexual parasite forms, the gametocytes, show a characteristic development over five distinct stages. The earliest stage is hardly distinguishable from mature schizonts. After stage II, the gametocytes grow into stage III forms with a similar size to schizonts. Stage IV parasites represent the largest of the gametocyte stages with a size of 10.9 µm (Dearnley et al., 2012). Stage V gametocytes are the most mature stage and are the only stage that is found in the bloodstream to eventually getting taken up by a mosquito. About 900 proteins are expressed in gametocytes, 315 of them are exclusive (Baker et al., 2011; Silvestrini et al., 2010).

Generally speaking, the IMC protein composition is very similar in asexual merozoites and presexual gametocytes. GAP50 forms a complex with GAP45 and MTIP and shows similar solubility profiles of the proteins (Dearnley et al., 2012). Even though the glideosome components are still present, the actin-myosin motor does not play a role in gametocyte elongation, but may be implicated in important structural tasks. In contrast to the architecture of the IMC in merozoites that appears to be formed of only one cistern (Kono et al., 2012), the IMC in gametocyte contains 10 – 15 cisternae that are connected at transverse sutures (Meszoely et al., 1987; Dearnley et al., 2012; Kono et al., 2012). Those sutures sub-divide the IMC into several plates, giving the parasite a segmented appearance. Moreover, they connect the IMC with the PM (Meszoely et al., 1987; Kono et al., 2012). Underneath the IMC lie the microtubules, which are spaced at intervals of approximately 10 nm (Dearnley et al., 2012).

(34)

1.4 Modifications of proteins

A variety of cellular functions is controlled by a range of different protein modifications. Small functional groups are often added to a specific amino acid residue, which is the case for one of the most common modifications, phosphorylation. Ubiquitination and nitrosylation are other well-studied modifications that regulate protein function. However, the attachments of lipid moieties to proteins like myristoyl, palmitoyl, stearoyl or farnesyl also play important roles in protein function and regulation. The most common forms of protein fatty acylations in eukaryotes are N-myristoylation and S-palmitoylation (Resh, 1999). The two fatty acylations can modify proteins both separately and concertedly with other lipid modifications. These modifications have become increasingly recognized to be of major importance for a better understanding of how sub-cellular localization, trafficking and enzymatic activity of membrane-associated proteins are regulated (Resh, 1999).

Figure 1.6 Chemical structure of palmitate and myristate. The enzyme-available form of both

palmitate and myristate consists of coenzyme A (CoA) linked to the lipid by replacing the hydrogen.

1.4.1 Myristoylation

Myristoylation is characterized by the co-translational addition of myristic acid, a 14-carbon fatty acid (Fig. 1.6). 0.5 % of all eukaryotic proteins are myristoylated (Maurer-Stroh et al., 2002). This irreversible process is catalyzed by cytosolic N-myristoyl transferases (NMTs) and can occur co- and post-translationally. Protein myristoylation is a well-understood process and a consensus motif has been identified as MGxxC/S/T (Resh et al., 1999; Maurer-Stroh et al., 2002). 10 to 14 carbons are inserted hydrophobically into the hydrocarbon core of the bilayer. Even though myristoylation is required for membrane association, it is not sufficient for stable and permanent membrane anchoring (Maurer-Stroh et al., 2002;

(35)

Aicart-Ramos et al., 2011). A second signal is needed for myristoylated proteins to become stably attached to a membrane. This can be a polybasic cluster or palmitate.

N-myristoyl transferase (NMT) mechanism and inhibition

There is little data on how myristoylation is regulated in vivo. NMTs can become phosphorylated by tyrosine kinases that, in turn, are myristoylated by NMTs (Selvakumar and Sharma, 2006). However, the mechanism of NMTs has been investigated in great detail and is composed of several steps: First, an amino peptidase cleaves off the initializing methionine. Myristoyl-CoA binds NMT with very high affinity and induces a conformational change so that the peptide substrate can bind. Nucleophilic substitution leads to attachment of myristoyl-CoA to the N-terminal glycine and a stable amide bond is formed. The myristoyl-CoA is then released, followed by the myristoylated peptide (Rudnick et al., 1991; Wright et al., 2010). The NMT of the fungus Candida albicans was crystallized already in 1998 (Weston et al., 1998) and shown to have internal twofold symmetry, the “NMT fold”. Its core domain is constructed predominantly of several α-helices (Weston et al., 1998).

Using myristic acid analogues, binding of different acyl chains has been investigated in great detail. The findings include that:

- chain length is important for binding of myristoyl-CoA to NMT (Heuckeroth et al.,

1988)

- functional and polar groups only have minor effects on NMT activity (Heuckeroth et

al., 1990; Devadas et al., 1992)

- myristoyl-CoA is bound in a bent conformation (Heuckeroth et al., 1988)

- the myristoyl-CoA binding site is highly conserved across species (Heuckeroth et al., 1988; Kishore et al., 1993)

- peptide recognition occurs within the first 10 N-terminal amino acids (Heuckeroth et al., 1988)

Maurer-Stroh and co-workers predicted a consensus motif by using the first 17 N-terminal amino acids of a protein to calculate whether a protein becomes myristoylated or not (Maurer-Stroh et al., 2002). Furthermore, they propositioned that there are three regions of a peptide that are needed for proper interaction with the enzyme: the very N-terminal region (residues 2 to 7) is bound by the active site, the central region of a peptide (residues 8 to 11) interact with

(36)

the surface opening of the catalytic pocket, and the last region (residues 12 to 18) act as a hydrophilic linker (Maurer-Stroh et al., 2002). NMTs have been characterized in several parasitic species, because the enzyme is essential to the survival within their host. Since the initial publication, crystal structures of the S. cerevisiae as well as P. falciparum NMT were revealed (Bhatnagar et al., 1998; Gunaratne et al., 2000). Several research groups have probed into selective inhibition of parasite NMTs (Panethymitaki et al., 2006; Bowyer et al., 2007; Bowyer et al., 2008; Frearson et al., 2010; Crowther et al., 2011; Tate et al., 2013). However, the high homology between human and parasitic NMTs and the conserved binding site for myristoyl-CoA complicates the process of finding specific inhibitors. Nonetheless, the protein substrates are different. This fact can be exploited to create pathogen-specific NMT inhibitors. N-myristoylation is vital to a cell’s survival. Especially promising was the discovery of an inhibitor of the NMT of Trypanosoma brucei, since it is among very few T. brucei proteins that were shown to be selectively inhibited (Frearson et al., 2010). Furthermore, a recent study showed that the plasmodial NMT is essential and can be targeted by various drugs (Wright et al., 2014). Compound 2a is a very promising candidate since it reduces parasitaemia in mice. Several substrates were identified as well, including IMC proteins like ISP1, ISP3, CDPK1, GAP45, MTIP, MyoA, alveolin 5. One phenotype of NMT inhibition is the failure of IMC assembly and failure to progress into merozoites (Wright et al., 2014).

1.4.2 Protein palmitoylation

Protein palmitoylation is the only reversible post-translational lipid modification known and it is characterized by the addition of the 16-carbon fatty acid palmitate (Fig. 1.6) to cysteine residues, forming a thioester linkage (Resh, 1999; Dietrich and Ungermann, 2004). As a consequence, palmitoylation is also called thioesterification. Palmitoylation is a regulatory mechanism that mediates protein-membrane attachment and sub-cellular trafficking of proteins. Furthermore, it plays an important role in protein-protein interactions, protein stability and enrichment of proteins in microdomains of membranes. Palmitoylation is the most common form of fatty acylation in eukaryotes and it is catalyzed by palmitoyl acyltransferases (PATs).

(37)

Palmitoyl acyltransferases (PATs)

Regardless of the discovery of palmitoylation several decades ago, the enzyme family of PATs, catalyzing this post-translational modification has only been discovered in recent times (Lobo et al., 2002; Roth et al., 2002). PATs reside in different tissues and sub-cellular localizations and usually act on intracellular proteins (Ohno et al., 2006; Mitchell et al., 2006; Batistic, 2012; Frénal et al., 2013). Some sequence requirements in PAT-substrates have been discovered and palmitoylation can be predicted using CSS-Palm (Ren et al., 2008; www.csspalm.biocuckoo.org). The prediction tools are based on a clustering and scoring strategy (CSS) algorithm. Through searching the scientific literature and collecting experimentally verified data on palmitoylation sites this tool was generated. 263 palmitoylation sites were analyzed and 4-, 6-, 8-, and 10-fold cross-validations and leave-one-out validation were calculated (Ren et al., 2008). Nevertheless, a clear general sequence motif has not yet been identified, although it has been shown that leucines, lysines and additional cysteines are favored amino acid residues surrounding palmitoylated cysteines (Lobo et al., 2002; Roth et al., 2002; Bartels et al., 1999; Babu et al., 2004; Smotrys and Linder, 2004; Roth et al., 2006; Xue et al., 2006; Ohno et al., 2006; Hou et al., 2009).

PATs are polytopic membrane proteins with four or more transmembrane domains (TMDs) and typically share a DHHC-motif (A: asparagine; H: histidine; C: cysteine), which is flanked by cysteine-rich domains (CRDs; 51 aa in length) and faces the cytosolic side of membranes, usually between TMD 2 and TMD 3 (Greaves et al., 2011; Mitchell et al., 2006). This motif is directly involved in the palmitoyl transfer and was found in all eukaryotic species examined so far. Particularly, H1 and C have been identified to be important for substrate binding (Mitchell et al., 2006). Experiments in yeast showed that interchange of DHHC motifs between different PATs cannot restore a PATs’ function (Mitchell et al., 2006). Therefore, the DHHC motif itself is required for substrate specificity, function and folding. Even though the general topology and the DHHC-CRD region seems to be rather conserved among PATs, they vary significantly at the sequence level (Batistic, 2012). Besides the DHHC-CRD motif, there is also a DPG (D: asparagine, P: proline; G: glycine) motif next to TMD2 and a TTxE (T: threonine; E: glutamate) motif next to TMD3, all of them facing the cytosolic side of the membrane. Whereas it is unclear whether or not those other domains have vital functions for enzyme activity, the C-terminal domains of PATs seem to be important for localization and substrate identification (González-Montoro et al., 2009; Beck et al., 2013). A second class of PATs exists; the membrane-bound O-acyl transferases (MBOATs) usually act on proteins that

(38)

are modified in the lumen of the secretory pathway. The number of PATs can vary from over 20 in metazoans (22 in Homo sapiens, 12 in P. falciparum and 18 in Toxoplasma gondii) to seven in the budding yeast Saccharomyces cerevisiae (Roth et al., 2006; Frénal et al., 2013).

Plasmodium falciparum expresses twelve proteins containing the DHHC motif (Jones et al.,

2012), of which six show distinct upregulation in late stages (Cabrera et al., 2012; www.plasmodb.org). Of note, using complementary palmitoyl protein purification approaches and quantitative mass spectrometry, over 400 palmitoylated proteins were identified in asexual blood stages of the parasite, including those involved in cytoadherence, drug resistance, signaling, development, and invasion (Gavin et al., 2002). Importantly, the physiological role of palmitoylation of these proteins is for the most part unknown and is not necessarily connected to membrane association. PATs reside in different sub-cellular localizations and might play a role in specific membrane recruitment of proteins such as VAC8 to vacuole membrane in yeast, RAS2 into the plasma membrane and ARO to the rhoptry membrane of T. gondii (Dietrich and Ungermann, 2004; Beck et al., 2013). Furthermore, it was reported that some PATs show dual localization patterns being localized to two organelles simultaneously (Ohno et al., 2006; Batistic, 2012). PATs are not only expressed in different cellular compartments, but may also be localized to different tissues (Ohno et al., 2006). Moreover, while some PATs are ubiquitously expressed, the expression pattern of others seems to be highly regulated.

A comprehensive localization map of PATs in T. gondii was recently established, where two of these PATs (TgDHHC2, TgDHHC14) could be localized to the IMC (Beck et al., 2013, Frénal et al., 2013). This study also identified two IMC-localized PATs in P. berghei (PbDHHC3, PbDHHC9), whereas only one out of 11 PATs has been localized to the rhoptries (PfDHHC7) in an over-expression approach (Frénal et al., 2013). Based on the fact that PATs localize to different organelles in Apicomplexa, it is tempting to speculate that PAT-substrate pairs exist. However, the most intriguing question is how substrate specificity of PATs is achieved.

Sequence specific recruitment of proteins to the inner membrane complex of the malaria parasite Plasmodium falciparum (Welch, 1897)